Aim: As the most biodiverse region of the world, the drivers of genetic diversity in Amazonia are still poorly understood. It has been debated that species on distinct ecosystems will show unique biogeographic histories tied to their habitats which in turn help understand landscape climatic history in Amazonia. We studied bird species associated with patchy Amazonian white-sand ecosystems (WSE) to evaluate the occurrence of shared patterns and its relationship to species habitat preferences and Amazonian environmental and landscape history.

Location: Northern South America; Amazonia.

Taxon: Passerine birds.

Methods: We sequenced Ultra-conserved Elements (UCEs) from 177 samples of seven bird species associated with WSE that have overlapping ranges. We used the SNP matrices and sequence data to estimate genetic structure patterns and migration surfaces using ‘conStruct’ and eems, perform model-selection to obtain the most probable demographic histories on ‘PipeMaster’, and implement analyses of shared demography with ecoevolity.

Results: Shallow genetic structure patterns varied among species with the Amazon river being the only barrier shared among them. Population structure dates to no more than 450,000 years ago. We identified 17 geographically structured populations from which nine showed signals of population size changes and eight of these occur in Northern Amazonia. Population expansion was inferred at two distinct times: ~100,000 years ago and ~50,000 years ago. The timing of co-expanding populations is consistent with differences in habitat preference, as species that prefer dense scrubby to forested vegetation expanded more recently compared to species that prefer open vegetation.

Main conclusions: We show that WSE species responded in concert to past environmental and landscape changes. Past population expansions were likely driven by the genesis of new WSE patches and a return to wetter conditions after glacial periods. Pleistocene climatic cycles affected the distribution and dynamics of open vegetation habitats in Amazonia, especially in the Northern region, driving genetic diversity and demographic patterns of its associated biota.

We used the phyluce v1.5.0 pipeline to process the UCEs from the raw reads (Faircloth, 2016; Faircloth et al., 2012). First, raw reads were cleaned from adapters and low quality bases using illumiprocessor and trimmomatic using phyluce v1.5.0 default parameters (Bolger et al., 2014; Faircloth et al., 2012). Next, the data was assembled into contigs using trinity and called from within the phyluce pipeline using phyluce_assembly_assemblo_trinity with default parameters (--KMER_SIZE set to 25) (Grabherr et al., 2011). Finally, we used the phyluce code to find, extract, and align (using mafft; Katoh & Standley 2013) the UCE loci from the assembled reads. This process was done separately for each species and generated species-specific alignments of every UCE locus.